Why animal charities are much more effective than human ones

Us­ing the chicken-to-hu­man re­sults on this SSC post (https://​​slat­estar­codex.com/​​2019/​​03/​​28/​​par­tial-re­trac­tion-of-post-on-an­i­mal-value-and-neu­ral-num­ber/​​) ( some of you may have seen it), I was in­ter­ested in just how much more effec­tive an ACE char­ity is over a pop­u­lar Givewell one, like Malaria Con­sor­tium. I re­ally do think that this an­i­mal vs hu­man de­bate should be talked about much more, con­sid­er­ing that (1) this is a very prac­ti­cal change we could all be mak­ing, and (2) it could make a huge differ­ence.

So the first as­sump­tion that has to be made here is that 25 chick­ens = 1 hu­man. Now in­tu­itively this prob­a­bly feels off to ev­ery­one that isn’t a hard­core an­i­mal welfare ad­vo­cate, even to me it still feels off, but let me try to ex­plain why it could make sense. Big­ger brains, like the hu­man one, gen­er­ally shouldn’t scale lin­early in moral worth be­cause of diminish­ing re­turns. The huge num­ber of cor­ti­cal neu­rons hu­mans have could just be go­ing to pro­cesses like think­ing, mem­ory, lan­guage, things that don’t con­tribute to the raw suffer­ing that is nec­es­sary for moral worth. Even in ne­ma­todes with 300 neu­rons, we still see them have an averse re­ac­tion to preda­tors, even the smell of them. This could mean that ne­ma­todes are ex­pe­rienc­ing some very prim­i­tive form of suffer­ing that gives them a lot more moral worth than imag­ined. Or as bet­ter ex­plained by this quote: “Neu­ral net­work analy­ses show that cog­ni­tive fea­tures found in in­sects, such as nu­meros­ity, at­ten­tion and cat­e­gori­sa­tion-like pro­cesses, may re­quire only very limited neu­ron num­bers. Thus, brain size may have less of a re­la­tion­ship with be­havi­oural reper­toire and cog­ni­tive ca­pac­ity than gen­er­ally as­sumed, prompt­ing the ques­tion of what large brains are for. Larger brains are, at least partly, a con­se­quence of larger neu­rons that are nec­es­sary in large an­i­mals due to ba­sic bio­phys­i­cal con­straints. They also con­tain greater repli­ca­tion of neu­ronal cir­cuits, adding pre­ci­sion to sen­sory pro­cesses, de­tail to per­cep­tion, more par­allel pro­cess­ing and en­larged stor­age ca­pac­ity. Yet, these ad­van­tages are un­likely to pro­duce the qual­i­ta­tive shifts in be­havi­our that are of­ten as­sumed to ac­com­pany in­creased brain size.”

Now the ques­tion is what chicken-to-hu­man num­ber do we use to com­pare the char­i­ties? The MTurk sur­vey that one of Scott’s read­ers used came up with 25, but what I use for com­par­i­son is a square root of neu­ron count, which is 15,000293,000 = 20 chick­ens per hu­man (chick­ens have 220 mil­lion and hu­mans have 86 billion to­tal.) This num­ber is eerily similar to 25, which is a bit com­fort­ing. How­ever this num­ber is only look­ing at ca­pac­ity to suffer. In or­der to com­pare this in du­ra­tion, we have to com­pare a broiler chicken life to a full un­der-5 life saved by Malaria Con­sor­tium. That would look some­thing like 6 weeks/​40 years = 330*20 = 6600 chick­ens-to-hu­mans. This num­ber feels a lot more in­tu­itively bet­ter now. The guessti­mate model (https://​​www.getguessti­mate.com/​​mod­els/​​10636) here of THL gives an es­ti­mate of 18 an­i­mals “spared” per dol­lar (this is re­ally just the 5% welfare im­prove­ment that equals a chicken). Malaria Con­sor­tium saves a life at around $2200. So the fi­nal calcu­la­tion here would be 6600/​​18 = 1 “hu­man” saved per $366. That’s about 6x more effec­tive than Con­sor­tium, and it’s only us­ing a 5% welfare im­prove­ment which sounds very low to me. So the call to ac­tion here would be to start pul­ling more dona­tions to THL or an­other ACE char­ity. Even us­ing very con­ser­va­tive es­ti­mates for welfare im­prove­ment, you still get a large in­her­ent differ­ence be­tween an­i­mals and hu­mans.